Motion interpolation by mean shift segmentation and motion analysis

We propose a novel motion interpolation algorithm by using the mean-shift segmentation and motion analysis technique. Mean shift segmentation is frequently used to extract objects from video according to its efficiency and robustness of non-rigid object tracking. For diminishing the computational complexity in motion estimation and object tracking process, an efficient block matching algorithm: cross-diamond-hexagonal search algorithm was used. In the motion analysis procedure, the stick figure of object obtained by thinning process is considered as guidance to gather the statistics of motion information. The model of new behavior of object is produced by motion analysis and used as an input to our motion interpolation procedure. The experimental results show that the proposed method changes the behavior of the behavior of object in several video sequences. Interested readers are welcome to visit our demo website at http://mineweb.no-ip.org/Umedia2008/.